West Virginia University
Sensor Placement Algorithms for Process Efficiency Maximization
Abstract
dc:description.abstractEven though the senor placement problem has been studied for process plants, it has been done for minimizing the number of sensors, minimizing the cost of the sensor network, maximizing the reliability, or minimizing the estimation errors. In the existing literature, no work has been reported on the development of a sensor network design (SND) algorithm for maximizing efficiency of the process. The SND problem for maximizing efficiency requires consideration of the closed-loop system, which is unlike the open-loop systems that have been considered in previous works. In addition, work on the SND problem for a large fossil energy plant such as an integrated gasification combined cycle (IGCC) power plant with CO2 capture is rare.;The objective of this research is to develop a SND algorithm for maximizing the plant performance using criteria such as efficiency in the case of an estimator-based control system. The developed algorithm will be particularly useful for sensor placement in IGCC plants at the grassroots level where the number, type, and location of sensors are yet to be identified. In addition, the same algorithm can be further enhanced for use in retrofits, where the objectives could be to upgrade (addition of more sensors) and relocate existing sensors to different locations. The algorithms are developed by considering the presence of an optimal Kalman Filter (KF) that is used to estimate the unmeasured and noisy measurements given the process model and a set of measured variables. The designed algorithms are able to determine the location and type of the sensors under constraints on budget and estimation accuracy. In this work, three SND algorithms are developed: (a) steady-state SND algorithm, (b) dynamic model-based SND algorithm, and (c) nonlinear model-based SND algorithm. These algorithms are implemented in an acid gas removal (AGR) unit as part of an IGCC power plant with CO2 capture. The AGR process involves extensive heat and mass integration and therefore, is very suitable for the study of the proposed algorithm in the presence of complex interactions between process variables.
Degree
thesis:*- Name thesis:degree_name
- PhD
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Chemical and Biomedical Engineering
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Paul, Prokash
- Contributors dc:contributor
-
- Debangsu Bhattacharyya
- Richard Turton
- Parviz Famouri
- Charter D Stinespring
- Stephen E Zitney
Identifiers
dc:identifier.*- Identifier
- https://researchrepository.wvu.edu/etd/6389